In this article I will explain why Advanced Micro Devices' (AMD) superiority in high-end graphics, which is the basis of the next era in computing, is more important than Intel's (INTC) current lead in the manufacture of semiconductors.
Process Lead has Not helped Intel in Mobile
Intel has been touting its advantage in shrinking the size of its processors (node), for over a decade now. Yet it has garnered less than 5% of mobile sales thus far. If a process lead is so important, why does Qualcomm (QCOM) dominate the mobile space using the inferior manufacturing of TSMC (TSM) and other Asian manufacturers? If a process lead is so important, why were Apple's (AAPL) mobile phones and tablets able to produce the most profits in the history of technology, without Intel inside?
For most of the last decade, Intel has been essentially saying, and I paraphrase, "just wait till next year, and we will have competitive mobile products." Every few years, Intel makes a move to a smaller manufacturing node, thereby maintaining its lead over the Asian contract Fabs, used by Qualcomm, AMD, etc. Each time this happens, most of the Intel fans proclaim, and I paraphrase again, "this new node will finally allow Intel to be competitive in mobile."
The Reality of the Contract Fab Era
Over the last decade or so, TSMC and Samsung (SSNLF) have increased their sales and profits, which has allowed them to decrease the capex gap with Intel. In a few years, TSMC and Samsung will be spending as much as Intel does on building new fabrication facilities for microprocessors. At that point in time, Intel will most likely lose its node lead over the Asian contract Fabs. Its not a matter of if, its a matter of when Intel loses its manufacturing lead over the contract Fabs. This situation is occurring because Intel failed to garner the design IP required to compete in mobile. The contract Fabs now have the economies of scale and the customers to compete with Intel.
Intel's factories rely upon the same equipment manufacturers as the Asian Fabs. In other words, the only real manufacturing advantage that Intel held was based upon its ability to outspend the contract Fabs.
When Intel no longer has a process advantage over the contract Fabs, it will be forced to compete with AMD based upon design strength/IP.
The fact that Intel's manufacturing lead has not helped in mobile thus far, is why I believe a similar situation will play out in the next era of computing, which is based upon the AMD's graphic superiority.
Q3 2013 will be the first quarter that AMD's graphic superiority manifests itself in rising sales and profits. This new trend is just beginning, and is solely based upon the gaming console space.
As time goes on AMD will be able to increase revenues and profits in other areas outside of the game console space. The most obvious area of growth, is the requirements of the data centers to provide video/graphics to thin client devices.
Constant gains in connectivity (bandwidth), allows many devices to utilize the huge power of cloud-based computing. Intel has dominated the server chip market with a 95% share, compared with AMD's about 5%.
As software is written to utilize the superiority and efficiency of AMD's high-end GPUs, rather than Intel's dominance in CPUs, the next era in computing will gain momentum.
The trends I have outlined here have just begun, and will assert themselves over a multi-year time frame.